ChefSense: An Intelligent Real-Time Cooking Quality Assessment System Leveraging Computer Vision, Advanced Feature Extraction, and Ensemble Machine Learning for Smartphone Applications

Abstract

Cooking quality assessment remains a critical challenge in food preparation, directly impacting nutritional preservation, food safety, consumer satisfaction, and culinary outcomes. Traditional evaluation methods depend on subjective human judgment, resulting in inconsistent assessments, significant inter-individual variability, and limited scalability for real-world applications. This paper introduces ChefSense, an innovative real-time cooking quality assessment system that integrates advanced computer vision techniques with ensemble machine learning algorithms to provide objective, automated food quality evaluation using standard smartphone cameras. Authors developed a comprehensive 208-dimensional feature extraction framework encompassing color distributions through RGB and HSV histograms, statistical properties including channel-wise means and standard deviations, and domain-specific indicators capturing brightness, dark ratios, browning characteristics, and texture density patterns. The system was rigorously evaluated using 7,720 food images from the Food-101 dataset across four distinct quality categories: perfect, good, overcooked, and burnt. Through systematic comparative analysis of ten diverse machine learning algorithms including Random Forest, Gradient Boosting, Support Vector Machines, Neural Networks, and probabilistic classifiers, our results demonstrate that Gradient Boosting and Decision Tree approaches achieve exceptional classification performance with 99.87% accuracy, 0.9987 precision, 0.9987 recall, and 0.9987 F1-score, substantially surpassing existing methodologies. The system exhibits remarkable computational efficiency with approximate 100ms processing latency per image, enabling practical real-time deployment on resource-constrained mobile devices without requiring specialized hardware or cloud connectivity. ChefSense represents a significant advancement in automated food quality monitoring, offering transformative applications in smart kitchen technologies, culinary education platforms, professional food service quality control, and consumer cooking assistance systems.